Font Size: 

Item format and presentation can impact how people respond to items on an instrument. Gender and age differences on positively and negatively worded items have been found with constructs such as anxiety and stress. In this study, two positive and eight negatively worded items from the Radloff (1977) CES-D depression scale were selected, with ten new items created to be reverse-worded matches for the original items in order to reduce the impact of construct differences versus wording differences. The purpose was to investigate the functioning of reverse-worded items measuring depression based on age and gender.

A volunteer sample of 2,540 adults was disaggregated by age and gender. POLY-SIBTEST was used to conduct differential bundle functioning analyses of ten negatively worded depression items using ten positively worded items as the matching subtest. When women were matched with men, women were significantly more likely than men to agree with negative items (b_uni = -.478, p = .021; ½ pt difference on a 4 pt scale). Comparisons by age category only were not significant. A follow-up analysis comparing males and females at three age categories revealed trends consistent with the global gender comparison; women more likely to agree with negatively worded depression items than men. However, the only significant difference was between males and females in the 40 to 59 year age category (b_uni = -0.852, p = .009). Although the difference between males and females 60 years and older was moderate, it was not significant (b_uni = -0.482, p = .092). There were no significant differences between males and females 18 to 39 years of age (b_uni = -0.216, p = .506). Results indicate that gender differences are consistent across age categories, however younger adults may be less impacted by positive versus negative item wording with depression items.